Use cases
Asset infrastructure for the job at hand.
How physics-ready content supports training, evaluation, synthetic data, and digital-twin workflows.
3D Assets for Better Sim-to-Real Transfer
Close the sim-to-real gap with validated OpenUSD and MJCF SimReady assets. Rigyd derives mass, friction and collision meshes and verifies them by test, so policies trained in simulation transfer to hardware.
View use case →
Simulations Across Every Scenario and Edge Case
Generate diverse, validated OpenUSD and MJCF assets for domain randomisation. Thousands of unique objects with physics verified by test, covering the variation a policy has to survive.
View use case →
3D to Digital Twin Pipeline
Turn 3D CAD and BIM models into physically accurate digital twins. Rigyd converts factory and warehouse models into SimReady OpenUSD with mass, friction, and collision meshes at thousands-of-objects scale.
View use case →
Physically Accurate Synthetic Data Generation
Generate physically accurate synthetic training data for robotics. Rigyd builds the SimReady 3D asset layer, mass, friction, collision meshes, that Isaac Sim Replicator, Omniverse, and custom pipelines need for transferable datasets.
View use case →
RL Training Environments
Build reinforcement-learning environments from validated OpenUSD and MJCF SimReady assets. Physics verified by test, collision geometry optimised so rollouts stay cheap, and an API and CLI to assemble the training world.
View use case →
SimReady Assets for Robot Policy Evaluation and Testing
Use Rigyd-generated SimReady assets for held-out policy evaluation, regression testing, safety scenarios, and pre-deployment validation. Reproducible, scalable, and matches the real environment.
View use case →